Cancer of the anal canal, a reality in the Colombian coffee region. Clinical-epidemiological review 2000–2019

Introduction Anal cancer is a rare pathology which has increased over the last few decades, and, therefore, gained importance for the quality of life of affected individuals. Thus, a review has been conducted in the Colombian coffee region (Departments of Caldas, Quindío y Risaralda) describing its behaviour and clinical-epidemiological profile. Materials and methods Descriptive review of 437 patients of Western SAS Oncologists between January 2000 and December 2019 with a diagnosis of anal cancer. Results 62% of cases presented in women with a median age of 62 years, 30% in the sixth decade; centred at 65% in three main cities designated as capitals (Manizales, Pereira and Armenia); 62% as localised disease, with 40% stage II-A and 6% as initial metastasis; 29% presented positive ganglia, particularly N1a; squamous cell or epidermoid histology in 90%; 16% poorly differentiated; 5% related to Human Immunodeficiency Virus infection; localisation in the medial area of the anal canal in 63% of cases; 83% completed treatment, and 92% of them received chemotherapy/radiation therapy with 87% based on the Nigro protocol; finally, 11% presented with relapse in the liver in 10% of cases and 55% local. Conclusion Four hundred and thirty-seven patients evaluated over 20 years with follow up at median 34.13 months (standard deviation 41.75) with median survival at later ages decreasing to 62% in patients older than 80 years, and differences in survival in localised disease at 78% in comparison to 46% in advanced metastasis. Finally, the overall 5-year survival rate is 69% with a median survival of 191 months in the study.


Introduction
Anal cancer is a rare pathology and is considered infrequent, accounting for 1%-2.6% of digestive cancers and 2%-4% of anal-colorectal tumours [1], with increased incidence worldwide [2] and an increasing rate of 2.9%/year since 1992, according to The Surveillance, Epidemiology, and End Results Program of National Cancer Institute (SEER)-USA [3]. In Europe, nearly 2,000 men and 2,300 women are diagnosed each year [4].
According to the Global Cancer Observatory (GLOBOCAN), an agency belonging to the International Agency for Research on Cancer of the World Health Organisation, the global incidence rate is >1.2 per 100,000 women and 0.6-0.89 per 100,000 men. In the USA, the rate is 1 per 100,000 men and 1.8 women, with an overall rate of 1.4 per 100,000 inhabitants and mortality of 0.2 per 100,000 inhabitants. In comparison, in South America, the incidence is 0.57 per 100,000 for men and 1.20 per 100,000 in women with an overall rate of 0.92 per 100,000 inhabitants and a mortality rate of 0.19 per 100,000 inhabitants [5].
In reviewing South American cases, references are found from Brazil-a retrospective of 25 years with 49 cases [6]; from Venezuela-a report of results from 96 cases from 2000 to 2013 [7], and a previous report of 56 patients, and finally, in Chile, the mortality rate was reported at a minimum of 0.11 in 2003, increasing to 0.21 per 100,000 in 2010 [8].
In Colombia, according to GLOBOCAN, an incidence of 557 new cases was reported (0.55%), occupying 26th place, and a mortality rate of 84 cases (0.18%) with a prevalence of 1,395 cases (prop.2.82) in 2018: for Colombian statistics, a 15-year review effected at the Bogota National Cancer Institute shows 18% of cases [10], and a report of the results of the analysis of 82 patients of Oliveros R et al [11] published in 1990, treated by the gastroenterology service at the Bogota National Cancer Institute from 1935 to 1981.
With respect to local statistics, according to the Manizales Quinquennial Population Cancer Registry 2003-2007, the incidence is 0.8% (12 cases) in men and 0.6% (11 cases) in women, with 0.7% overall. During the 5-year period from 2008 to 2012, it was 0.5%, with an incidence rate of 0.9 per 100,000 inhabitants [12] and more specifically, in the area of the Caldas Department, a frequency of 0.7% is found (26th place) [13].
Anal cancer is an uncommon neoplasm; however, its incidence has recently increased due to Human Immunodeficiency Virus (HIV), making it the fourth most common cancer in this population with a general annual rate of 1 per 100,000 inhabitants to 35 per 100,000 inhabitants in HIV(+) men. Other factors include anal sex, promiscuity, immunosuppression, effects of Human Papillomavirus (HPV) attributed to HPV-16 serotype, the cause in both genders of 73% of cases, with the HPV-18 serotype attributed to 3.4%-7% [14]. According to the Centers for Disease Control and Prevention, these main serotypes [15] are present in 79% of anal cancers, attributed to the use and abuse of cigarettes, which causes an increase in cases in both sexes with worldwide predominance in males in this population [16]. The Colombian Coffee Sector (Caldas, Quindío and Risaralda Departments) does not escape this reality due to the scarcity of epidemiological data. In view of this, an analysis of the effects of anal cancer in the Western SAS Oncologists' area of influence was performed to provide evidence of the magnitude of the problem, its clinical profile, epidemiology, treatment and finally, the results attained for this pathology.

Materials and methods
A retrospective analysis was performed from the database of the Western SAS Oncologists, a private organisation dedicated to cancer awareness, treatment and monitoring in this geographical area, from 01/01/2000 to 31/12/2019, using International Disease Classification codes 10th edition (CIE-10) with C21.0, C21.1, C21.2 and C21.8 related to anal cancer [17]. The CIE-10 was employed as a base classification as the institutional standard for the classification of cancer patients who were examined, and with compromised histology as an exclusion criterion, thus raising doubts about anatomical location in the final selection of a population base for the analysis of 437 cases.
To begin, general clinical and demographic aspects were analysed (age, occupation, procedure, department, municipality, clinical stage, oncological classification by tumour, nodules and metastasis (approved by the 2017 investigatory team at tumour-node-metastasis (TNM) to avoid confusion and time variation), HIV, Karnofsky and symptom duration); anatomical variables (histology, histological grade, prognostic factors, anatomical location within the canal); therapeutic variables (surgery, type of non-surgical treatment, gap in radiation therapy, initial and rescue chemotherapy plans, radiation therapy dose and sessions, site of recurrence, time at recurrence, number of rescue chemotherapy treatments and a description of the therapy as completed or not) and finally, based on the results of intervention and survival (overall monitoring and survival according to different variables analysed). These factors may be analysed and defined as risk factors according to the results obtained. The date of diagnosis was analysed in 5-year increments to evaluate the patterns, age by 10-and 20-year period and the clinical stage as defined by the most recent classification of the American Joint Committee on Cancer (AJCC) from 2017 [18].
The histological grade was classified as well differentiated, moderately differentiated and poorly differentiated, defined by an institutional or external pathology team, and was based on the pathological findings following established guidelines for the speciality, which is outside the scope of this analysis. The treatment was based on surgery, radiation treatment, chemotherapy, radiation and chemotherapy together or none. Recurrence, or clinical evidence of tumour activity, was evaluated by the oncologist or, when in doubt clinically, based on a positive biopsy report confirming tumour activity following treatment, or without the necessity of a pathological study if a lesion, defined by paraclinical methods such as endoscopy or radiology, was present, with all of the foregoing occurring following a disease-free period of 12 months or more. The persistence of the tumour was defined as aforementioned, without a disease-free period, after completing the principle treatment. The gap variable was described as the time, in days, of radiation therapy interruption, with a goal of permitting functional recovery of the patient from generally morbidity from chemotherapy/radiation therapy [6]. Symptom duration is given in months.
Monitoring was defined as the period from the date of diagnosis to the final completed treatment, and survival was defined as the period from the end of treatment to remission or death. The final stage was determined using categories of alive or dead, sub-classified as active, lost and deceased, with or without tumour activity.
Qualitative variables were analysed as proportions, and quantitative variables are calculated using standard deviations. Survival calculations were performed using the Kaplan-Meier estimator and the log-Rank test. Statistical analysis was based on Epi Info 6.04™, Version 3.

Results
Anal cancer constitutes 2% of colorectal anal tumours, presenting twice as often in females (62%) and tripling in frequency beginning in the fifth decade (Table 1), with survival diminishing as age increases, from 75% in the 26-40-year-old age group to 62% for ≥80 years, displaying an inverse relationship between age and survival (p = 0.37) ( Figure 1).
87% of the group who completed treatment received the Nigro protocol, in comparison to 6% receiving the Mayo protocol, 4% receiving Capecitabine/Mitomicina C (MMC) as a variant of the Nigro protocol and 3% receiving 5-Fluoropirimidina (5-FU)/Cisplatin (CDDP), with no statistically significant differences between Nigro, 5-FU/CDDP and Mayo ( 5 OS: 75% versus 68% versus 62%, p = 0.34), though favouring the alternate protocol of Capecitabine/MMC in comparison to Nigro (91% versus 75%, p = 0.58). Comparing the protocols designated as second line, with overall rates of response nearing 70% and mOS of 12 months for 5-FU/Cisplatin, slightly less than Nigro ( Figure 4). As with other adverse factors, status as described by the Karnofsky scale becomes, in the majority of neoplasms, a further prognostic factor.
The total dose for radiation therapy in the group who completed treatment was 49.08 Gys, with differences within the group of greater or less than 45.00 Gys, favouring the larger applied dose ( 5 OS: 77% versus 68%, p = 0.01) ( Table 3). Dx; diagnostic, Frec; frequency  32% presented with recurrence, with a statistically significant difference in survival between persistence (17%) and recurrence (47%) (p = 0.00), with this difference being larger upon comparison between patients who did or did not present with recurrence (36% versus 90%, p = 0.00), along with differences between patients who completed treatment in comparison to those who did not: 52% in 5 OS and a mean survival (mOS) of 177 months, favouring the group who completed treatment, to ultimately generate an overall survival of 95%, 89%, 85%, 81% and 69% from the first to the fifth year, and an mOS of 191 months (Table 4).
Regarding the most common clinical stage, II, presenting in 54%, differences are found in 5 OS of 12% between the II-A and II-B sub-types (p = 0.02), as also found by others [1][2][3], showing that the differences were essentially in tumour size (smaller or larger than 5 cm), defined in the 2017 AJCC and establishing the TNM II-B sub-type for future analysis as a risk factor for localised stages [20,22,24]. Data for 5 OS for stage Zero-I was 100%, with 54%, 38% and 48% for stages III-A, III-B and III-C, with mOS of 78, 42, 50 and 18 months for stages III-A, III-B, III-C and IV, respectively, and, finally, not reaching mOS in localised stages (Zero, I, II-A and II-B) (p = 0.00), showing once again the TNM T4 sub-type to be an adverse prognostic factor in comparison to ganglion compromise in tumours of smaller size II-A (T1-T2 with positive ganglia) [20,32].
Ulceration presented as an adverse factor in 5 OS in 56%, similar to that reported by others, and constituting an adverse factor for consideration in future investigations, alongside node status (p = 0.00) and male gender (0.01) [33].
In the 6% of cases representing initial metastatic disease, the liver was the most affected site at 50%, similar to others, with data showing between 5% and 10% for metastatic disease [7,19,23,29,34] constituting an independent adverse prognostic factor with maximum survival of 28% in comparison to 75% for the group without metastasis, and mOS of 28 versus 191 months (p = 0.00), respectively, with differences resulting from the comparison of local disease with locally advanced/metastatic disease, such as the majority of the series wherein a regimen other than the Nigro protocol was proposed [35][36][37][38].
Regarding HIV and its role in the development of anal cancer, 1% of women and 28% of men with anal cancer have an underlying infection, with this neoplasm occurring more in this population in comparison with the general population. In 4.6%, HIV was found to be present; a low number, but maintaining the relationship to male gender, similar to others [28,29,39,40], with marked numerical differences between HIV (+) and HIV (−) populations, favouring the negative ( 5 OS: 9% versus 66%, p = 0.08), without the ability to compare the larger series in which they were excluded [30,32,33,41], and with contradictory results in different studies depending upon whether they were included in the analysed population [14,42].
Similar to other authors, functional status is shown as a prognostic factor with differences in mOS between patients with Karnofsky ≤ 40%, 50%-70% and ≥80% of 8 and 180 months (10, 18 and 190 months, p = 0.00), respectively ( Figure 5) [19,41]. 89% completed treatment based on concomitant chemotherapy/radiation therapy, of whom 87% received the Nigro protocol [35], generating differences upon grouping them as chemotherapy/radiation therapy, other treatments (surgery, radiation therapy, chemotherapy) and no treatment (p = 0.00). In the group who completed treatment, the differences between patients treated with Nigro (87%) compared to 5-FU/Cisplatin and others were 5 OS: 72%, 62% and 56% (p = 0.10). This further confirms that expressed by multiple authors since the Nigro protocol was first presented at the Meeting of the American Proctologic Society in Detroit in June of 1973, and published in 1974 [35]. The improvement with the Nigro protocol, which has become the standard treatment, shows minimal differences in comparison to 5FU/CDDP according to reports by others [29,31,37,38], and corroborated by our results from the Nigro protocol versus 5-FU/CDDP ( 5 OS 71% versus 48%, p = 0.04) and a difference of 151 months in mOS favouring the Nigro protocol (mOS 191 versus 40 months) [31,32,37,38].
Analysing the other protocols employed, 5% received the Mayo protocol, with the difference ( 5 OS: 55% versus 71%, p = 0.06) in favour of Nigro, and 4% received an alternative protocol, Capecitabine/MMC, with results showing slight benefit to the alternative protocol ( 5 OS: 91% versus 75%, p = 0.58), with no statistically significant difference in concordance with different authors based on non-inferior results, such as alternative therapies in patients who were not candidates for Nigro [23,29,43,44] for medical reasons, co-morbidities or lack of availability of infusion pumps to administer the Nigro protocol.
Radiation therapy, specifically for the group who completed treatment (76%), gave differences in favour of a higher dose, 45.00 Gys ( 5 OS: 68% versus 77%, p = 0.01). Analysis of different applied doses shows, similar to others, that better survival is not reflective of increased dose [45], as is demonstrated by </> 60.00 Gys, with the difference favouring the group receiving the lower dose ( 5 OS: 38% and mOS: 140 months, p = 0.57), defining the optimal dose for localised stages as between 45.00 and 50.40 Gys, and possibly for more advanced stages according to TNM T and N(+), a larger dose is required, up to 54.00-60.00 Gys [19]. With respect to a gap during radiation therapy, this presented in 24%, leading to differences of 2% in 5 OS and 11 months in mOS, favouring the use of a gap in a non-statistically significant manner (p = 0.33), contrary to others who conclude that poorer outcomes are reflective of a gap [34,46] (Figure 6).  Recurrence was noted in 11% at an average of 39.83 months (standard deviation 35, 39); 55% at a local level and 10% in the liver, similar to what others have noted [47,48]. Persistence in 21% generated differences in 5 OS between the presence of persistence, recurrence and no recurrence of 12%, 67% and 84%, respectively, (p = 0.00), and, upon grouping the data as recurrence or no recurrence, the differences remain (40% versus 84%, p = 0.00). Regarding rescue therapy, similar to others, the practice of only making a therapeutic attempt leads to differences of 30% in 5 OS and 140 months in mOS (p = 0.13), in comparison with not attempting it. Finally, with respect to rescue chemotherapy, the protocols are based on 5-FU/CDDP (F-FU/Cisplatin), followed by Nigro (5-FU/Mitomycin) the overall population, favouring 5-FU/CDDP as the rescue. In modern series with few patients, triconjugate protocols are employed; others are based on taxanes, platins, and, finally, based on targeted and immunomodulatory therapy, without reaching an ideal rescue protocol to replace 5-FU/CDDP [36,37,43,[47][48][49], as shown in (Figure 7), with 82% of the patients categorised as completing treatment and classifying them on the basis of whether or not recurrence occurred, 32% presented with this event, with 53% receiving rescue treatment and completing it in 96% of cases.

Conclusion
Anal cancer is increasing in the Colombian Coffee Area, with localised disease comprising the majority of cases, and the highest probabilities of cure ultimately defined by an epidemiological profile of women between 50 and 70 years of age, homemakers, from urban areas, initial stage of II-A, being T2, without ganglia and without initial metastasis. In the case of presenting ganglia of type N1a, with well-differentiated epidermoid histology and ulceration, 5% with HIV, the localised lesion found in the medial portion of the anal canal, a good functional state represented by Karnofsky ≥ 80% and symptoms over 6-12 months, who could receive treatment with chemotherapy/radiation therapy on the Nigro protocol with no gap, with a radiation therapy dose greater than 45.00 Gys. Despite the foregoing, recurrence at the locoregional and/or hepatic level occurred in 32% after 3 years of monitoring, and was principally treated with surgery, radiation therapy or chemotherapy/radiation therapy using the 5-FU/CDDP protocol reaching an average monitoring period of 39.42 months to ultimately generate an overall survival of 95%, 89%, 85%, 81% and 69% from the first through fifth years of follow-up. For the aforementioned, it was considered that the work complied with the objectives formed from the start to show the behaviour of this neoplasm in the coffee region.

Conflicts of interest
The authors declare no conflicts of interest of any kind in the collection, analysis of information or the results from an academic, labour or economic point of view.

Funding statement
The concepts presented herein are the sole responsibility of the authors, and the work was entirely financed by the investigators' own resources without the intervention of any company or entity.